Expression of the Chemokines MIP-1α, MCP-1, and RANTES in Experimental Autoimmune Uveitis

Abstract

purpose. To determine the location of the CC chemokines macrophage inflammatory protein (MIP)-1α, monocyte chemoattractant protein (MCP)-1, and regulated on activation of normal T-cell–expressed and secreted (RANTES) during disease progression in experimental autoimmune uveitis (EAU) and their relationship with the presence of the T helper cell (Th)1–type cytokine IFNγ.

methods. EAU was induced by immunization of Lewis rats with retinal extract. Consecutive cryostat sections were prepared from eyes at different stages of EAU, graded for severity of uveitis and stained by using antibodies to MCP-1, MIP-1α, and RANTES and to cell surface markers. Supernatants from superficial cervical lymph node cells were examined by ELISA for IFNγ, IL-4, and IL-10.

results. MIP-1α and IFNγ were present most frequently and most extensively at peak disease but also were detectable in the choroid 8 days after immunization, before clinical disease onset. MCP-1 and RANTES were present at peak disease, but much less frequently. RANTES was occasionally found in the choroid before clinical disease. By days 19 to 21 after immunization, although infiltrating cells were present, there were only residual low levels of chemokine staining. MCP-1 and RANTES were detected on CD3-positive cells and on some ED1-positive cells, whereas MIP-1α was also associated with vessels and areas of exudate. Lymph node cells cultured from animals with peak disease had increased levels of IFNγ and IL-10, but for IFNγ this occurred only after stimulation in vitro with retinal extract.

conclusions. Although MCP-1 and RANTES were associated predominantly with cells infiltrating the retina, MIP-1α was also associated with resident cells. All three are likely to exacerbate EAU—MIP-1α, to the greatest degree.

Experimental autoimmune uveitis (EAU) is an animal model for the inflammatory eye disease endogenous posterior uveoretinitis (EPU), which is thought to have an autoimmune origin. Immunization of animals at distant sites with retinal antigens and appropriate adjuvants results in a disease with many of the clinical and histopathologic features of the human disease. ¹ In EAU and EPU, the blood–retina barrier is breached, allowing leukocytes, particularly monocytes and antigen-specific and nonspecific T lymphocytes, to move into the retina, causing tissue damage and leading to impaired vision and in some cases blindness. ² Macrophages are thought to act primarily as effector cells, responsible for much of the tissue damage, with T cells in a regulatory capacity. ³  

Chemokines, which act as chemoattractants and activators of specific leukocytes at the site of inflammation may be involved in this influx of inflammatory cells. ⁴ Chemokines have been shown to be likely to be involved in leukocyte recruitment in the eye in a study of patients in which IL-8, IP-10, monocyte chemoattractant protein (MCP)-1, regulated on activation of normal T-cell–expressed and secreted (RANTES), and macrophage inflammatory protein (MIP)-1β were significantly increased in the aqueous humor during the active stages of anterior uveitis. ⁵ The chemokine family can be divided into two major classes, CC and CXC, on the basis of differences in the positions of cysteines within a conserved four-cysteine motif. ⁶ The CC chemokines, such as MIP-1α, MIP-1β, RANTES, and MCP-1, -2, and -3 have powerful chemoattractant and activator properties for monocytes and T cells, ^{4 6} whereas the CXC chemokines, such as IL-8 and those that, like IL-8, contain the ELR motif, are particularly important in the attraction of neutrophils. ⁶ Because evidence suggests that the predominant cells in EAU are macrophages and T lymphocytes, not neutrophils, ¹ we examined the location of the CC chemokines MIP-1α, MCP-1, and RANTES during disease progression in EAU. 

Helper T (Th) lymphocytes may be categorized in terms of the cytokines they produce, with Th1 cells producing IFNγ and IL-2 and Th2 cells producing IL-4, IL-5, IL-6, and IL-13. ⁷ Although IL-10 in humans is produced by both Th1 and Th2 subsets, in rodents it is classified as a Th2-type cytokine. ⁸ It is thought that Th1-type T cells contribute to the pathogenesis of organ-specific autoimmune diseases. ⁹ In patients with clinically active and inactive uveitis, there is evidence of a Th1-type response. ^{10 11} Th1-type T cells in experimental autoimmune encephalomyelitis (EAE), which models the human demyelinating disease multiple sclerosis, ¹² and in insulin-dependent diabetes mellitus ¹³ have been shown to have the ability to transfer disease. EAU has also been shown to be mediated by Th1 cells, ¹⁴ with susceptibility to EAU related to ability to generate a Th1-type response. ^{15 16 17}  

Chemokine production has been shown in some instances to be linked to Th1 response. For example, in one EAE model, production of both MIP-1α and RANTES could be correlated with Th1 responses and MCP-1 with Th2 responses. ¹⁸ Th cell subtypes have also been shown to carry receptors for different sets of chemokines, Th1 cells with receptors for MIP-1α, and RANTES. ¹⁹ We therefore examined the production of IFNγ alongside that of the chemokines to determine how chemokine location during disease development in EAU is related to a Th1 response. Location of the chemokines rather than site of synthesis is likely to be important, because it is possible that chemokines are presented and act at sites away from those at which they have been produced—for example, production by inflammatory cells or cells within the tissue and then immobilization and presentation on the luminal surface of endothelial cells by proteoglycans. ²⁰  

Methods

Induction of EAU

Female Lewis rats (Harlan Olac, Bicester, UK), 6 to 8 weeks old, were used throughout the study and were treated humanely according to the UK Animal License Act and to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. EAU was induced with a single subcutaneous injection into one hind footpad of 100 μl containing 0.4 mg retinal extract ²¹ emulsified 1:1 (vol/vol) in complete Freund’s adjuvant (CFA) containing 1 mg/ml Mycobacterium tuberculosis H37 Ra (Difco Laboratories, Detroit, MI). Control groups included animals injected with CFA and PBS in place of retinal extract and unimmunized normal Lewis rats. Animals were examined by slit lamp microscopy and clinical signs of ocular inflammation scored on a scale of 0 to 4.5. ²²  

Immunohistochemistry

After the animals’ death under terminal anesthesia using CO₂, eyes were embedded in optimal temperature cutting compound (OCT, Tissue-Tek; BDH, Poole, UK), snap frozen in isopentane-liquid N₂ and stored at −30°C. Consecutive 8-μm cryostat sections were prepared on poly l-lysine-coated slides (BDH) and air dried overnight. For detection of chemokines, the tissue sections were fixed in formaldehyde (2%) for 20 minutes at 4°C and washed three times in buffer containing saponin (Tris-buffered saline [TBS ]-0.1% white saponin[ BDH] ²³ ). This buffer was used for the remainder of the procedure, unless stated otherwise. Sections were incubated with 10% swine serum in buffer containing avidin B blocking solution according to the manufacturer’s protocol (Vector, Burlingame, CA) for 15 minutes before overnight incubation with primary antibody (1:100) in buffer containing biotin blocking solution (Vector). Primary antibodies to rat MCP-1, MIP-1α, and RANTES were polyclonal rabbit antibodies obtained from PeproTech EC, Ltd. (London, UK). Control antibody was rabbit IgG (PeproTech EC). IL-4 antibody was a goat anti-rat antibody (R&D Systems Europe, Ltd., Abingdon, UK) and control goat IgG (R&D Systems Europe, Ltd.), and sections were preincubated with 10% rabbit serum. Sections were washed three times in buffer for 5 minutes with rocking before incubation for 30 minutes with 1:100 biotinylated swine anti-rabbit antibody (Dako, Ltd., Cambridge, UK), which had been preabsorbed with 10% heat-inactivated normal rat serum for 30 minutes (1:100 biotinylated rabbit anti-goat antibody for IL-4). After they were washed, sections were incubated with streptavidin, biotinylated alkaline phosphatase complex (Dako) for 30 minutes, according to kit protocol before addition of fast red substrate solution. 

For detection of cell surface marker proteins, sections were first fixed in acetone for 10 minutes, air dried, and incubated in TBS for 30 minutes with the following primary monoclonal antibodies: OX6 (1:50), ED1 (1:100), ED2 (1:100), R73 (1:50) and W3/25 (1:50), all from Serotec (Oxford, UK). OX6 detects major histocompatibility complex (MHC) class II molecules: ED1 and ED2, infiltrating blood-borne macrophages, and resident tissue macrophages, respectively: R73 (T-cell αβ receptor) and W3/25 (CD4). Isotype control IgG (Dako) was used at the same concentrations. Sections were washed three times in TBS before incubation for 30 minutes with biotinylated rabbit anti-mouse antibody (1:100, Dako), which had been preabsorbed with 10% normal rat serum for 30 minutes. Signal was amplified and detected as for chemokine detection. For IFNγ detection, sections were fixed in acetone as described earlier and incubated with primary monoclonal antibody to IFNγ (1:100, Serotec). The procedure was continued as described, but the buffer used contained saponin. 

All incubations were at room temperature and all chemicals were obtained from Sigma (Poole, UK), unless stated otherwise. 

Sections stained for cell surface markers were used to grade the disease and obtain a semiquantitative assessment of severity of uveitis according to both structural–morphologic and infiltrative changes, by using a slightly modified customized version of the grading system established in this laboratory. ²⁴ Sections stained for chemokines were also assessed semiquantitatively in terms of the percentage area of the retina stained compared with an IgG control antibody. This was performed by observers masked to the stage of disease and antibodies used. 

Dual-Immunofluorescence Microscopy

Sections obtained from animals 11 days after immunization and demonstrating peak EAU were prepared, fixed, and blocked according to the procedure for chemokines and incubated with the first set of antibodies, CD3, ED1, IFNγ, MCP-1, and control IgG all 1:100 for 45 minutes. Buffer throughout the procedure contained saponin (TBS-0.1% white saponin). Sections were then washed three times for 5 minutes with rocking. Appropriate biotinylated secondary antibody was preabsorbed with 10% normal rat serum for 30 minutes before incubation with the sections for 45 minutes and washing as before. Sections were then incubated with 1:50 streptavidin Texas red (Amersham Life Science, Ltd., Buckinghamshire, UK) for 45 minutes in the dark. Sections were washed, and this and all subsequent procedures were performed with light excluded. The second set of antibodies, anti-rat RANTES, MCP-1, and MIP-1α at 2 mg/ml, were labeled using a protein-labeling kit (Alexa 488; Molecular Probes, Eugene, OR) according to the manufacturer’s instructions. They were diluted to 1:25 and preincubated with 10% normal rabbit serum for 30 minutes before addition to the sections and overnight incubation. Slides were washed three times for 30 minutes each wash with rocking. All incubations were at room temperature unless stated otherwise. Immunofluorescent staining was visualized using a laser scanning confocal imaging system (model 1024; Bio-Rad Laboratories, Ltd., Hemel Hempstead, UK). Retinal cells that were labeled fluorescently were counted in each high-power field. 

Culture of Lymph Node Cells

Superficial cervical lymph nodes were disrupted by passing them through a 250-μm stainless steel sieve. After two washes in RPMI, cells were resuspended at 5 × 10⁵/ml in RPMI with 4 mM l-glutamine, 100 IU/ml penicillin, 100μ g/ml streptomycin, and 10% fetal bovine serum (FBS) and cultured for 72 hours, with or without retinal extract at 0.8 mg/ml. Supernatants were harvested by centrifugation at 300g for 10 minutes and stored at −80°C until ELISA. 

Enzyme-Linked Immunosorbent Assay

Immunoreactive cytokine was quantified by sandwich ELISA for rat IFNγ, IL-4, and IL-10 (Biosource International, Camarillo, CA), according to the manufacturer’s instructions. Supernatants were added in duplicate, and the cytokine being tested was revealed with a monoclonal antibody conjugated to horseradish peroxidase. The minimum amount of cytokine detectable using these assays was less than 13 pg/ml for IFNγ, less than 2 pg/ml for IL-4, and less than 5 pg/ml for IL-10. 

Statistical Analysis

Experiments were repeated a minimum of three times, with four animals in each experimental group. Data are presented as the mean ± SEM. The statistical significance of the results was assessed using Student’s unpaired two-tailed t-test (GraphPad Software, Inc., San Diego, CA). 

Results

Immunohistochemical Detection of MIP-1α, MCP-1, RANTES, and IFNγ

On day 5 after immunization neither clinical nor histologic grading identified any changes associated with uveitis, and this was also the case for all the control animals whether they were examined on day 5 after immunization or at later stages after immunization. None of the chemokines could be detected in the retina, either at day 5 after immunization (Table 1) or in the control groups throughout the study. On day 8 after immunization (Table 1) , although there were no structural changes, activated cells could be detected mainly in the choroid and, rarely, in the retina. Occasional staining for MIP-1α (Fig. 1B) , IFNγ, and, very rarely, RANTES (Fig. 1D) could be detected in the choroid. IFNγ staining could be detected infrequently (Fig. 1A) in the retina. 

By days 11 to 12 disease was severe, graded clinically at 4 to 4.5, and immunohistochemically there was marked invasion of the photoreceptor layer by infiltrating cells that were strongly OX6 positive—both T cells (W3/25-positive and R73-positive) and ED1-positive macrophages (data not shown). There were loss of rod outer segments, retinal folds and exudative detachment of the retina, granuloma formation, and retinal vasculitis and perivasculitis. MIP-1α and IFNγ were detected strongly at this stage, associated predominantly with infiltrating cells and with granuloma formation and vasculitis (Table 1 , Figs. 1F 1G ). MCP-1 and RANTES were also detected at this stage of EAU, although less frequently than MIP-1α and IFNγ (Table 1 ), and associated with more discrete patches of infiltrating cells (Figs. 1H 1I) . IL-4 was not seen (Fig. 1J) . MIP-1α and MCP-1 were sometimes detected associated with choroidal and retinal vessels, as well as infiltrating cells; however, dual-immunofluorescence staining indicated that in most cases they did not localize to the same cells (Fig. 2A) . At this stage staining for IFNγ, MIP-1α, MCP-1, and RANTES was also apparent in the ciliary body and iris. 

At day 21 after immunization, there was extensive retinal destruction, and disease remained active. Although the inflammatory cell infiltrate was less, it was still strongly OX6 positive, with many W3/25-positive T cells and ED1-positive macrophages (data not shown). MIP-1α, MCP-1, and IFNγ were detected, but to a much lesser extent than at peak disease (Table 1) . RANTES, however, was not detected. 

To identify more clearly which cells were producing the majority of the chemokines in EAU, dual-immunofluorescence microscopy was performed on sections showing peak EAU. More ED1-positive cells were detected in the infiltrate (>75%) than CD3-positive cells; however, whereas most of the CD3-positive cells produced chemokines, only a small percentage of the ED1-positive cells did (Table 2A) . MCP-1 and RANTES were detected on CD3-positive cells mainly and on some ED1-positive cells (Table 2B) . Unlike MCP-1 and RANTES expression, which was predominantly associated with infiltrating inflammatory cells, substantial MIP-1α expression was also seen associated with the exudate in the subretinal space (Fig. 2B) and with vessels (Figs. 2C 2D) . 

Cytokine Production by Draining Lymph Node Cells

Levels of IFNγ and IL-10 above background were not detected in cultures of lymph node cells from control animals or immunized animals showing no signs of EAU either by clinical or histologic grading (day 5). However, cells cultured from animals with peak disease (day 12), as determined clinically and histologically, had increased levels of both IFNγ and IL-10 (Fig. 3) . In the case of IFNγ, this increase was only apparent after stimulation of the cultures with retinal extract, indicating that increased IFNγ production was a result of antigen-specific T cells present in the cultures and that a Th1-type response was occurring. This increase in IFNγ production was significant (P < 0.05) when compared with cultures, whether unstimulated with retinal extract or from undiseased animals or animals at day 21. IL-10 was significantly (P < 0.05) increased in diseased animals compared with undiseased or those at day 21, whether or not the cultures were stimulated with retinal extract, indicating that IL-10 production was not a function of antigen-specific T cells. By day 21, both IFNγ and IL-10 had decreased to levels comparable with those of undiseased animals. IL-4 was not detected in any of the samples (data not shown). 

Discussion

This study examined for the first time the location of the chemokines MIP-1α, MCP-1, and RANTES as the autoimmune inflammatory disease EAU developed. Of the three chemokines investigated, MIP-1α appeared to be most closely associated with the development of EAU. It was present most frequently and most extensively, particularly at peak disease, but also was detectable in the choroid 8 days after immunization, before clinical disease onset. 

The location of chemokines during EAU was examined, rather than their site of production. The use of saponin, ²³ however, locates chemokines within a cell as well as on the cell surface. Dual-immunofluorescence staining indicated that MIP-1α, unlike MCP-1 and RANTES, was not only associated with infiltrating cells but also with endothelial cells and other non-CD3-positive and non-ED1-positive cells, which may include microglia. In EAE, astrocytes produce MIP-1α. ^{25 26} MIP-1α, in addition to its detection, specifically associated with cells, was also shown to be associated with acellular exudates in the subretinal space. These acellular exudates are a result of lysis and fragmentation of the photoreceptors and vessel leakage during peak disease, and areas of fibrin deposition may be present. ¹ The ability of MIP-1α to bind to proteoglycans ²⁷ may account for its association with these exudates. 

In EAE, mRNA expression for a variety of chemokines including MIP-1α, MIP-1β, RANTES, and MCP-1 has been detected before the onset of clinical disease and throughout acute clinical disease. ²⁸ As with our findings in EAU, MIP-1α expression appeared to be of particular significance in the pathogenesis of some EAE models. In EAE induced by transfer of proteolipid protein (PLP)-specific T cells into SJL/J mice, antibodies to MIP-1α, but not to MCP-1 or RANTES, could inhibit the development of acute and relapsing EAE, as well as the infiltration of mononuclear cells into the central nervous system (CNS). ^{18 25} However, using DNA vaccines in a model in which Lewis rats were immunized with myelin basic protein (MBP), it was shown that vaccines for MIP-1α or MCP-1 prevented EAE, whereas RANTES vaccine had no effect. ²⁹  

Further evidence for a role for MCP-1 in autoimmune inflammatory disease has been provided by the MBP model in which EAE and anterior uveitis develops in Lewis rats. In this model MCP-1 was detected preclinically in the iris-ciliary body and lumbar spinal cord, increasing as disease developed and coinciding with expression of IL-2 and IFNγ. ³⁰ It was suggested that MCP-1 contributed to the initial recruitment of inflammatory cells into both the eye and CNS. However, in our study MIP-1α rather than MCP-1 appeared more likely to be important in early recruitment, because it was detected before MCP-1, at day 8 after immunization. This difference may be linked to the less acute nature of posterior uveitis. 

Both MIP-1α and MCP-1 were detected in association with choroidal and retinal vessels, both with endothelial and perivascular cells. However, dual-immunofluorescence microscopy indicated that they did not localize to the same cells. Similar findings have been described at the blood–brain barrier, where separate binding domains for MIP-1α and MCP-1 have been identified on the parenchymal surface of brain microvessels. ³¹  

IFNγ detection in the tissue sections paralleled disease development and also MIP-1α production. As with MIP-1α, IFNγ was also detectable in the choroid at day 8 after immunization, extensive at peak disease, and much reduced by days 19 to 21 after immunization. This supports earlier studies on IFNγ mRNA expression in EAU. ^{32 33} IFNγ is produced by T cells, but dual-immunofluorescence staining clearly showed that MIP-1α was produced by other cells in addition to T cells. IFNγ production by T cells may be stimulating MIP-1α production by macrophages. In contrast, in rheumatoid arthritis T cells from synovial fluid express MIP-1β and RANTES in the absence of IL-2 and IFN-γ, and the investigators speculate that these chemokines downregulate the expression of specific T-cell–secreted cytokines, such as IL-2 and IFN-γ, which can alter the Th1-to-Th2 balance. ³⁴ There is no evidence to suggest that this happens in EAU. 

The involvement of IFNγ in EAU indicated by immunohistochemistry was corroborated by the production of IFNγ by cells cultured from the draining lymph nodes, which was increased at peak disease but only by antigen-specific cells. IL-4 was not detected, but IL-10 production was increased at peak disease; however, unlike the increase in IFNγ, this increase occurred both with and without antigen stimulation. IL-10 can indicate a Th2 response, but IL-10 is also produced by macrophages and is therefore not as definitive a marker of Th subtype as is IFNγ. It is likely in this situation, given the substantial macrophage presence, that IL-10 is more indicative of macrophage activation. 

By day 21 after immunization, both IFNγ and IL-10 production by draining lymph node cells had returned to background levels, although infiltrating cells were still present in the retina, providing no evidence that an increased Th2 response is important for disease resolution and indicating, on the contrary, that a decrease in Th1 response is more significant in this respect. The absence of a Th2 cytokine increase in spontaneous remission of EAE ³⁵ and studies on the basis for genetic susceptibility to EAU also indicate that an inhibited Th1 response may be more important than an elevated Th2 response. ^{15 16} The importance of IFNγ in rat EAU has previously been shown in transgenic animals in which constitutive ocular expression of IFNγ led to early onset and increased severity of EAU. ³⁶  

In conclusion, the presence of the chemokines MIP-1α, MCP-1, and RANTES in the choroid and retina during the course of EAU is likely to aid the cellular infiltration of both antigen-specific T cells and effector macrophages. These chemokines may not be involved in the actual initiation of cellular infiltration, as they were not detected before infiltrating cells. MIP-1α, in particular, was found concomitant with the Th1 cytokine IFNγ, but this may be coincidental rather than causal, in that both are products of activated infiltrating cells. As there is no evidence for MIP-1α before IFNγ, MIP-1α may not alter the Th1-to-Th2 balance by attracting Th1 cells over Th2. T cells may produce MIP-1α and also stimulate macrophages to produce MIP1α through IFNγ. Production of MIP-1α may then result in a cascade effect by promoting the entry of more monocytes and T lymphocytes into the retina. 

 Supported by The Wellcome Trust Grant 049531.

 Submitted for publication July 5, 2000; revised January 3, 2001; accepted January 18, 2001.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Detection of Chemokines and IFNγ in Posterior Segment after Immunization with Retinal Extract

Days after Immunization	Grading*	MIP-1α	MCP-1	RANTES	IFNγ
5 (n = 4)	0/1	−	−	−	−
8 (n = 4)	0/1	+	−	+	+
11 (n = 10)	4/4	++++	++	+	+++
19–21 (n = 6)	3/3	+	+	−	+

 Chemokines and IFNγ were detected by immunohistochemical staining, which was graded semiquantitatively as the percentage area of the retina stained compared with an IgG control section. −, No staining; +, <1%; ++, 1%–10%; +++, 10%–25%; ++++, 25%–50% stained. Data are from three experiments with n the total number of animals examined. ²⁴

*  Structural/infiltrative: Disease was graded in terms of structural changes and infiltration of leukocytes into the retina using cell surface markers. Highest score possible 5/6.²⁴

Figure 1.

 

Immunohistochemical staining of cryostat sections from rats immunized with retinal extract and CFA. (A–E) Eight days after immunization: (A) IFNγ, (B) MIP-1α, (C) MCP-1, (D) RANTES, and (E) IL-4. (F– J) Eleven days after immunization, showing retinal detachment, choroidal inflammation, and inflammatory cells: (F) IFNγ, (G) MIP-1α, (H) MCP-1, (I) RANTES, and (J) IL-4. (K, L) Control sections 11 days after immunization: (K) CFA only, stained for MIP-1α, and (L) IgG control. ILM, inner limiting membrane; INL, inner nuclear layer; ONL, outer nuclear layer; ROS, photoreceptor rod outer segment layer; ch, choroid, INF, infiltrating inflammatory cells; SRS, subretinal space underlying detached retina. Magnification, (A–E) ×125; (F, G) ×100; (H, J) ×125; (I)× 200; (K, L) ×65.

Figure 2.

 

Dual-immunofluorescence staining of cryostat sections from rats immunized with retinal extract and CFA at 11 days after immunization. Section stained with (A) antibody to MIP-1α conjugated to a fluorescent dye (green) and antibody to MCP-1 stained with Texas red showing infiltrating cells in subretinal space underlying detached retina; (B) antibody to MIP-1α conjugated to a fluorescent dye (green) and antibody to ED1 stained with Texas red showing infiltrating cells and exudate (arrow) in subretinal space underlying detached retina; (C, D) antibody to MIP-1α conjugated to fluorescent dye (green) and antibody to ED1 stained with Texas red showing retinal vessels in inner nuclear layer. Bar, (A) 37 μm; (B–D) 50μ m.

Table 2.

 

Staining Positivity

A. Percentage of ED1- and CD3-Positive Cells Also Staining Positively for MIP-1α, MCP-1, and RANTES
	MIP-1α	MCP-1	RANTES
ED1	7.8 ± 4.5	3.2 ± 1.8	8.1 ± 2.2
CD3	92.9 ± 7.1	100 ± 0.0	100 ± 0.0

B. Percentage of Chemokine-Positive Cells Also Staining Positively for Either CD3 or ED1
	CD3	ED1	Non-CD3 or ED1-Positive Cells
MIP-1α	25.0 ± 17.9	7.1 ± 2.0	67.9
MCP-1	78.5 ± 21.4	12.0 ± 7.2	9.5
RANTES	76.1 ± 9.5	36.9 ± 20.2	0

 Dual immunofluorescence microscopy on tissue sections showing peak EAU. Experiments were repeated three times. Data are ±SEM.

Figure 3.

 

IFNγ and IL-10 in supernatants of lymph node cells cultured with or without retinal extract (RE) for 72 hours. Results are from a representative experiment with four animals at each point. Error bars indicate the SEM. −RE, without retinal extract; +RE, with retinal extract.